Method and system for identifying person crossing a belt based on computer vision

Abstract

The invention discloses a method and a system for identifying a person crossing a belt based on computer vision, and the method comprises the steps: obtaining a person and a belt in a video through a trained instance segmentation model; fitting curves of two edge contours of the belt; determining the foot position of the person; and judging whether the person crosses the belt or not according to the position relation between the foot position and the curves of the two edge contours of the belt. An image of a belt area is collected through an existing camera, key information features of a belt and personnel are extracted through an instance segmentation algorithm, a judgment algorithm for the personnel to cross the belt is set by combining the geometric position relation of the personnel and the belt, and real-time and efficient judgment is achieved. Compared with monitoring by ordinary people, a large amount of manpower is saved, and real-time monitoring is guaranteed. Moreover, the implementation of the invention only needs to install the camera and deploy the server on site, thereby reducing the management complexity.

Description

technical field [0001] The invention belongs to the fields of computer vision and underground personnel construction safety, and in particular relates to a computer vision-based identification method and system for personnel crossing a belt. Background technique [0002] With the increasing attention to the safety of energy mining, energy companies should pay more attention to the personal safety of on-site workers while focusing on increasing production capacity. For some dangerous behaviors, early detection and early treatment should be carried out to avoid personal injury. [0003] At present, the method of ensuring the safety of personnel in underground mine sites is still relatively traditional, such as installing cameras on the job site for remote monitoring by a special person, or setting up on-site security personnel, etc., which require a lot of manpower and are prone to human error Unsafe behaviors have not been discovered due to the interference of various factors...

AI Technical Summary

Problems solved by technology

[0003] At present, the method of ensuring the safety of personnel in underground mine sites is still relatively traditional, such as installing cameras on the job site for remote monitoring by a special person, or setting up on-site security personnel, etc., which require a lot of manpower and are prone to human error Unsafe behaviors have not been discovered due to the interference of various factors, resulting in certain hidden safety hazards

Images